Latest / Mechanical Engineering Made Simple / (#77) Combined Stress Deflection Energy Matrix Algebra

Transcript

(#77) Combined Stress Deflection Energy Matrix Algebra · Mechanical Engineering Made Simple

• 0:00
  Welcome back design junkies to Mechanical Engineering Made
• 0:02
  Simple. Today we were taking a deep dive
• 0:05
  into the very core of structural analysis, tackling the concepts
• 0:09
  that well that separates a simple calculation from a
• 0:12
  reliable, complex structural design.
• 0:15
  Our mission is to demystify 2 massive areas engineers wrestle
• 0:18
  with everyday. First, we need to understand how
• 0:22
  real world loads combine to create complex stresses.
• 0:25
  You know we're talking about components that are being both
• 0:28
  pushed and bent at the same time.
• 0:30
  And 2nd, we'll explore the powerful methods we use.
• 0:32
  We'll trace the evolution from classical difficult integration
• 0:37
  to the elegant shortcuts of energy principles, and finally
• 0:40
  the numerical power of matrix algebra.
• 0:42
  All to predict if a component will deflect too much or just
• 0:45
  fail out, right? It's a perfect map for this
• 0:47
  journey. We are definitely stepping
• 0:49
  beyond the easy case of, you know, pure tension or
• 0:51
  compression. Structural success, especially
• 0:53
  when you're dealing with non uniform loads at all, hinges on
• 0:56
  accurately modeling those combined effects.
• 0:58
  And as soon as you have those combined stresses, you
• 1:01
  absolutely must have rigorous, accurate deflection analysis.
• 1:05
  That need for efficiency is what pushed engineers toward well
• 1:10
  sophisticated tools like energy methods and the vast matrix
• 1:13
  algebra that drives every modern simulation today.
• 1:16
  OK, let's unpack this starting where the rubber meets the road
• 1:20
  combined stress. As you said, it's rare that a
• 1:22
  real column or a strut only feels forced directly down at
• 1:26
  center. So if I have a short stocky
• 1:27
  strut, maybe steel post, and the load is slightly off center,
• 1:31
  what exactly happens inside that material?
• 1:34
  What happens is a beautiful demonstration of superposition
• 1:37
  when that component is subjected to simultaneous bending and
• 1:40
  thrust, the resultant longitudinal stress.
• 1:43
  We call it Sigma. It's really just the sum of 2
• 1:45
  perfectly separate stress systems.
• 1:47
  So you're just adding two different effects together.
• 1:49
  Exactly. You're combining two types of
• 1:51
  deformation. The first and the simplest 1 is
• 1:54
  the average compressive stress from the axial load P That's
• 1:57
  just P / A where A is the cross-sectional area.
• 1:59
  That's your uniform squeeze. Got it.
• 2:01
  That part's straightforward. It's evenly distributed across
• 2:03
  the whole area. But the second part is the non
• 2:06
  uniform stress, the one created by the bending moment M.
• 2:09
  That stress is calculated as m * y and then divided by I sub X.
• 2:14
  And remind us what Y and I sub X are.
• 2:17
  Sure. So Y is the distance for the
• 2:18
  neutral axis and I sub X is the second moment of area which is.
• 2:23
  It's basically that shapes inherent resistance to being
• 2:26
  bent. It's the stiffness that comes
• 2:27
  from the geometry itself. So if we're analyzing a point,
• 2:30
  we literally just calculate the uniform squeeze and then we add
• 2:34
  or subtract the stress caused by the bending at that exact point.
• 2:37
  Sigma equals P / a + m Y over I sub X.
• 2:42
  Exactly right. And that brings us immediately
• 2:44
  to a critical real world scenario.
• 2:47
  Eccentric thrust. When the load is off center.
• 2:49
  Right when the load P acts slightly off center, we call
• 2:52
  that distance the eccentricity E sub Y.
• 2:54
  It is statically and mathematically equivalent to two
• 2:56
  simultaneous effects. Acting right at the center.
• 2:58
  You have a simple axial thrust P + a pure bending moment M where
• 3:02
  m = P multiplied by E sub Y. OK, so this is what you'd see
• 3:06
  in, say, poorly aligned machinery or columns where the
• 3:11
  connecting beam isn't perfectly centered.
• 3:13
  Precisely the offset distance, no matter how small, it
• 3:17
  automatically converts part of that downward force into a side
• 3:20
  to side rotational moment, and that dramatically changes the
• 3:24
  stress profile. Which leads to some very
• 3:26
  important design rules, I imagine.
• 3:28
  It leads to a crucial one, especially in materials like
• 3:31
  masonry or concrete, which have practically no strength and
• 3:34
  tension. We have the no tension rule.
• 3:36
  When you combine that bending and thrust tensile stresses, the
• 3:40
  pulling stresses are going to occur in the fibers farthest
• 3:44
  from the center, on the side, away from the load.
• 3:46
  So you're squashing one side, but actually pulling the other
• 3:48
  side apart. That's it.
• 3:50
  And for a component with the simple rectangular cross
• 3:52
  section, if you want to ensure that no part of the material is
• 3:55
  being pulled apart, I mean 0 tensile stress, the eccentricity
• 4:00
  E sub Y must be limited. The load P can only act within
• 4:04
  the middle 1/3 of the section. The absolute limit on that
• 4:07
  eccentricity is 1/6 of H where H is the depth of the column.
• 4:10
  If the load moves beyond that central core, that 1/6 limit,
• 4:15
  you start creating internal tension.
• 4:17
  And in concrete, that means cracking, and that means
• 4:20
  eventual failure. That one sixth rule is a
• 4:22
  fantastic, memorable detail. It just shows that in structural
• 4:25
  engineering, a tiny deviation and load placement can
• 4:29
  completely change the failure mechanism.
• 4:32
  Now let's pivot from strength, whether it breaks, to stiffness,
• 4:35
  which is how much it moves. Why is deflection just as
• 4:39
  critical as strength and design? Well, stiffness is non
• 4:42
  negotiable. It's because structural
• 4:44
  components have to be both strong and stiff enough.
• 4:46
  A strong beam that sags too much might not collapse, but.
• 4:50
  But it's still a failure. It's a functional failure,
• 4:52
  Absolutely. You can crack ceilings, cause
• 4:54
  equipment misalignment, or just be unusable.
• 4:57
  The design has failed. So mathematically, how do we
• 5:00
  capture that inherent resistance to bending?
• 5:03
  We use the term flexural stiffness, which you'll always
• 5:05
  see is EE, is Young's modulus, the materials intrinsic
• 5:08
  stiffness and I is that second moment of area we just talked
• 5:12
  about the geometric stiffness. So a stiff design has a high
• 5:15
  EIA. Very high EI and the fundamental
• 5:19
  relationship for elastic bending links the curvature, which is 1
• 5:22
  / R directly to the loading M In the stiffness it's 1 / R = m
• 5:27
  divided by EI. This symbol relationship though
• 5:30
  immediately takes us down the rabbit hole of classical
• 5:33
  calculus. It leads to the governing
• 5:35
  differential equation for deflection.
• 5:36
  OK, lay on us. If V is the vertical deflection
• 5:39
  along the length Z, the equation is EI multiplied by the 2nd
• 5:43
  derivative of V with respect to Z ^2 = m, the bending moment.
• 5:48
  That equation looks simple enough on paper, but solving it
• 5:51
  in the real world, I mean for a complex beam, maybe one with a
• 5:54
  distributed load that stops halfway or point loads all over,
• 5:58
  that requires integrating that moment equation multiple times
• 6:01
  and dealing with those discontinuities sounds like a
• 6:02
  nightmare. It absolutely is.
• 6:04
  When the load changes abruptly, the equation for M changes and
• 6:08
  you have to perform multiple integrations applying boundary
• 6:10
  conditions every single time. It just it turns into pages of
• 6:13
  messy work. So how did engineers find an
• 6:15
  elegant fix for this mathematical headache?
• 6:18
  Well, necessity is the mother of invention, right?
• 6:20
  That need to handle discontinuities efficiently led
• 6:23
  to tools like Macaulay Brackets. The square.
• 6:26
  Brackets. Those specific square brackets.
• 6:27
  They simplify the whole integration process.
• 6:30
  The clever part is that the term inside the brackets only
• 6:34
  applies. It only exists mathematically
• 6:37
  when that term is positive. So if a load starts at Z = 4 and
• 6:41
  your bracket is Z -, 4, that load only counts when Z is
• 6:45
  greater than 4. You've got it.
• 6:47
  It's an elegant fix to avoid integrating piecewise functions
• 6:51
  over and over. It packages all the transition
• 6:53
  points into one unified equation.
• 6:56
  But even that is still calculus. If we're trying to find the
• 6:58
  deflection on a massive, statically indeterminate
• 7:01
  structure, one where you can't even find the support forces
• 7:04
  with simple statics, the calculus must get overwhelmingly
• 7:07
  complex. It does.
• 7:08
  So what was the next great conceptual breakthrough?
• 7:11
  The shortcut to the whole deflection problem.
• 7:13
  That's where we make the conceptual leap to energy
• 7:15
  methods. For statically indeterminate
• 7:18
  structures, where you have more unknown reactions than equations
• 7:21
  of equilibrium, classical integration is almost
• 7:25
  impossible. Energy methods just bypass the
• 7:28
  entire force calculation and go straight to displacements.
• 7:30
  OK. Tell us more about the
• 7:31
  foundation of this. The foundation is the principle
• 7:34
  of virtual work. Simply put, for any system in
• 7:38
  equilibrium, the network done by the forces during any arbitrary
• 7:42
  virtual displacement is 0. A virtual displacement, so a
• 7:46
  tiny imaginary movement. Exactly.
• 7:48
  When we apply this to beams, we equate the external work done by
• 7:52
  a unit load, a single reference force.
• 7:54
  We place where we want the displacement to the internal
• 7:57
  work stored as strain energy. The equation becomes the
• 8:00
  integral over the length L of m * 1 over RDZ, which equals the
• 8:05
  work done by our unit load, let's say one moving through the
• 8:08
  actual displacement delta. So you're equating external work
• 8:11
  to internal energy. That's beautiful, but it still
• 8:14
  sounds like we have an integral to solve.
• 8:15
  Where's the powerful shortcut? The truly powerful nugget is the
• 8:18
  concept of complementary energy, which we call C.
• 8:21
  This is the ultimate engineering shortcut for finding deflection.
• 8:25
  Instead of integrating moments, you just define the total
• 8:29
  complementary energy function for the whole structure.
• 8:31
  And here's the key. The magic is in differentiation,
• 8:35
  not integration. The partial derivative of that
• 8:38
  complementary energy function C with respect to an external
• 8:42
  load, say W sub one, gives you the corresponding displacement
• 8:46
  delta sub one. Wait, really?
• 8:48
  Really. You plug in your total energy
• 8:49
  function, take a derivative with respect to the load you care
• 8:52
  about, and out pops the displacement at that exact
• 8:55
  location. It's a direct, elegant path to
• 8:58
  the answer. Wow, if classical calculus is
• 9:01
  mapping a journey mile by mile, energy methods are just looking
• 9:04
  at the start and end points to find the displacement.
• 9:06
  That conceptual groundwork I mean.
• 9:08
  It must have paved the way for modern computation.
• 9:10
  Absolutely. The energy principles, coupled
• 9:12
  with methods like Raleigh Ritz formed the critical foundation
• 9:15
  for all modern numerical analysis.
• 9:17
  They showed engineers how to turn structural behavior into a
• 9:20
  system of solvable linear equations.
• 9:23
  Leading us directly to the numerical realm, which is
• 9:26
  necessary because when you jump from a simple column to an
• 9:29
  aircraft fuselage or an offshore oil rig, the complexity scales
• 9:33
  beyond anything manual calculation can handle.
• 9:36
  Exactly when the structure becomes that massive and
• 9:39
  interconnected, manual methods breakdown completely.
• 9:42
  We have to transition to numerical methods, and the
• 9:44
  fundamental language they speak is matrix algebra.
• 9:48
  OK, for listeners who might be a bit rusty on this, let's keep it
• 9:50
  grounded in this context. What role does a matrix actually
• 9:54
  play? A matrix is simply a structured
• 9:56
  way to store and manipulate massive amounts of data at the
• 9:59
  same time. Just think of it as a
• 10:01
  rectangular table of numbers in structural mechanics.
• 10:04
  It lets us represent connectivity, material,
• 10:06
  properties, forces, all in an organized, computable way.
• 10:10
  So. It's about organization.
• 10:11
  It is. Knowing a matrix is, say, order
• 10:14
  4 by 3-4 rows and three columns tells a computer exactly how
• 10:19
  much data it needs to process and how it relates.
• 10:21
  And why are these arrays so vital?
• 10:23
  What core principle do they allow us to implement?
• 10:25
  Matrices are the foundation of the matrix displacement method,
• 10:28
  and from that the finite element method or FEM.
• 10:31
  They let us represent the entire complex structure as a single
• 10:35
  vast system of simultaneous linear equations.
• 10:38
  The system is often expressed very compactly as a relationship
• 10:41
  between 3 vectors. The force vector P which has all
• 10:44
  the external loads, the stiffness matrix K which holds
• 10:47
  all the properties in geometry. The DNA of the structure you
• 10:50
  could say. That's a great way to put it.
• 10:51
  And the displacement vector U which contains all the unknown
• 10:55
  deflections. The relationship is simple P = K
• 10:58
  * U. So if we can solve for those
• 11:00
  unknown displacements, you know exactly how the structure is
• 11:03
  behaving under load P and the finite element method fefem.
• 11:07
  That's the ultimate manifestation of this matrix
• 11:09
  approach, isn't it? It is.
• 11:11
  FEM takes that initial complex structure and using that matrix
• 11:14
  framework, subdivides it into thousands or millions of simpler
• 11:18
  finite elements like little triangles or bricks.
• 11:21
  The behavior of each small element is calculated with
• 11:23
  matrix algebra, and then all those individual stiffness
• 11:26
  matrices are compiled into a massive global stiffness matrix
• 11:30
  K By solving P = K * U for the whole system, we get the nodal
• 11:35
  displacements and all the resulting stresses everywhere.
• 11:37
  It lets us simulate the complex deformation of an entire
• 11:40
  suspension bridge or the wing of an airplane.
• 11:43
  That's right. That is an incredible evolution.
• 11:46
  We began with combining simple axial stress with bending stress
• 11:49
  to handle eccentric thrust, a problem solved with a formula
• 11:53
  and a simple rule like the 1/6 limit.
• 11:55
  Then we wrestled with the painful calculus for classical
• 11:58
  deflection before gratefully accepting the elegant shortcut
• 12:01
  of energy and virtual work methods.
• 12:03
  And finally, we see how that groundwork leads directly to the
• 12:06
  powerful numerical framework of matrix algebra and FEM, letting
• 12:10
  us tackle any structural complexity imaginable with a
• 12:12
  computer. And the utility of mastering
• 12:15
  these concepts is profound. Whether you're calculating the
• 12:18
  strength of a component by hand or simulating a structure on a
• 12:21
  supercomputer. The theory anchors the safety
• 12:24
  factors. For instance, the source
• 12:27
  material emphasizes the load factor, the amount loads can be
• 12:30
  increased before total collapse. Understanding those theoretical
• 12:33
  limits which are based on these core stress and stiffness
• 12:36
  concepts is just, yeah, it's paramount to ensuring a safe
• 12:40
  design. That theoretical grounding is
• 12:43
  everything, So building on our earlier discussion, we learned
• 12:47
  that in composite bars like steel and concrete, stress
• 12:50
  adjusts based on the Young's module IVE values to ensure they
• 12:53
  stretch equally. But here's something to think
• 12:55
  about. If you design A steel reinforced
• 12:58
  concrete beam where the materials have vastly different
• 13:01
  elastic properties, what unexpected internal stresses
• 13:04
  arise before any external load is even applied, simply due to
• 13:07
  construction or temperature changes?
• 13:09
  That's something for you to Mull over as you consider how those
• 13:12
  initial internal forces dictate a structure's final capacity.
• 13:16
  Thank you for joining us for this deep dive into the world of
• 13:18
  mechanical analysis. Always a pleasure exploring the
• 13:21
  structural scaffolding of the world.